DX Platform

What do you like best about the product?

How it merges quantitative and qualitiative feedback, and is rooted in data.

What do you dislike about the product?

I think it works fine for me - I don't have any immediate grievances.

What problems is the product solving and how is that benefiting you?

Figuring out what the bottlenecks in our engineering org are.

What do you like best about the product?

Useful insights on what is really going on at team level and beyond.

What do you dislike about the product?

I like all the reports but its a bit difficult to consume and find the right one you are looking for.

What problems is the product solving and how is that benefiting you?

Visualizing the right metrics (e.g. DORA) which I can tie in with my flow metrics reporting which gives a very good view for upper management.

What do you like best about the product?

The framework as a whole provides engineering managers with useful insights about how their teams is/are performing, taking into account several variables. Most of such insights are presented in easy to follow reports and charts and should be used by a team as metrics on which to improve.

What do you dislike about the product?

There is a significant overlap between the aspects that make the framework/platform appealing and the things that make it very dangerous: metrics are great, but will cause more harm than good when used used by people without context.

The framework allows people to compare metrics taken by their face value and without any real context. These comparisons can happen among people inside a team, among different teams and between a team and the industry.

For as much as the owners of the framework/platform say that such comparisons should not be made, they are inevitable - the framework/platform leads itself to such comparisons easily.

It's only a matter of time until somebody will say "this person is not contributing as much as others", "this team is not delivering on pair with other teams", "this team is severely behind what industry expects" and use this information to make poor decisions.

This gets even worse if we consider that the "industry" includes all kinds of companies that deliver software, regardless of them delivering a single use case product or a whole framework that allows customers to build any product.

As an example, any team that mostly does maintenance work, in which it is fairly common to have someone add 2-3 lines of code per PR after an investigation of several days to understand the root cause of the problem, will have significantly lower metrics than a team who is focused on delivering features. A team who works on a monolithic code base with several years of age will have significantly lower metrics than a team who only has to build and maintain the code for a new service delivered 2 years ago.

Such comparisons should be disabled by default. At the very least, these comparisons should only be available at a team level, so that only the people with right context can use such metrics wisely.

Another thing that I dislike about the framework is the AI magic behind some of the metrics. A given metric is defined as "AI weighs each PR based on PR title and description, the size and structure of the diff (files changed, lines changed, etc.), and other structural metadata about the PR.". Another metric is defined as "PRs are weighed and classified by DX AI". All of this is very vague, even with the provided examples, and not really useful for people to know how to improve. The good news is that is vague for everybody...

The framework/platform would largely benefit from replacing these vague descriptions with concrete fact-based rules.

While one can find some documentation online that tries to explain some of the concepts and metrics, this documentation is not comprehensive and sometimes it is just as vague as the metrics' description itself.

What problems is the product solving and how is that benefiting you?

The framework allows me to understand how my team is doing regarding several factors and understand how we can become better.

What do you like best about the product?

DX brings all relevant engineering metrics into one place. I mainly use it to track throughput (especially TrueThroughput) and delivery metrics, and correlate them with AI adoption. Data Studio is also very useful for building custom dashboards and using the AI chatbot to explore specific questions.

What do you dislike about the product?

There are not many issues overall. One limitation we recently encountered is the restriction around using a single Slack workspace. In our case, we have a subsidiary with its own engineering organization and separate Slack workspace, and we were not able to find a way to integrate them into the same DX tenant.

What problems is the product solving and how is that benefiting you?

DX helps me lead a nearly 100-person engineering organization with better data. It gives me visibility into staffing, AI adoption, throughput, and team-specific bottlenecks, which helps me make more informed decisions during a period of rapid change.

What do you like best about the product?

There are many metrics and I especially liked the AI adoption reports. The Claude Code adoption report was the only way I could get a sense of how much the members of my team are using Agentic AI.

What do you dislike about the product?

Same as the pro: it's a lot. Some metrics are not important for me.

What problems is the product solving and how is that benefiting you?

I am more informed about my team's health and overall engagement.